A STYLE OF MUSIC CHARACTERIZED BY

FIBONACCI AND THE GOLDEN RATIO

CASEY MONGOVEN
Abstract. Many composers of the 20th century used Fibonacci
numbers and the golden ratio in their works. None of these
composers, however, made them the basis of a style. Creat-
ing a style of music characterized by Fibonacci numbers and
the golden ratio requires discarding musical traditions and re-
thinking stylistic elements from the ground up. In this style,
mathematical properties of sequences related to the golden ratio
and Fibonacci numbers are converted into musical properties.
This paper introduces the style, presenting three short works
based on Fibonacci-related sequences.
The purpose of this paper is to introduce a style of music which has
its roots in the Fibonacci numbers and the golden ratio. Although
many composers of the past used these mathematical objects in their
compositions, none made it the basis of their musical style. Instead,
these composers superimposed the Fibonacci numbers and the golden
ratio on top of their own style in most cases in order to provide a
robust formal structure.
1. Short History of Fibonacci in Music
The rst references to the golden ratio in connection with mu-
sic appear to stem from German philosopher Adolf Zeising [1], who
referred to the proportion in 1856 as:
that standard proportion which underlies the ar-
rangement of the human form, the construction of the
more beautiful animals, the construction of plants,
namely in the form of their leaf-arrangement, the
forms of various crystals, the arrangement of the plan-
ets, the proportions of architectural and sculptural
works recognized as being the most beautiful, the
1
2 CASEY MONGOVEN
most satisfying chords in musical harmony, as well
as many other things in nature and art.
Although Zeisings hypothesis aroused great interest in the eld
of aesthetics, it was confronted with great skepticism by the scien-
tic community. This is evident in the writings of Gustav Theodor
Fechner [2], who wrote in 1870:
Because he [Adolf Zeising] is of the opinion that the
same principle of pleasantness must be valid in both
the eld of vision and sound and does not want to
hear of anything at all that deprives golden ratio of
power he nds himself obligated, in clear contradic-
tion to his general judgment, to declare both [major
and minor] sixths to be the most musically pleasant
proportions, for which he believes to nd reasons in
musical compositional proportions. As a result, he
would clearly be forced instead of the pure sixth, to
declare the impure sixth which corresponds exactly
to the golden ratio, to be the musically most pleasant
proportion. . .
One can see the widespread inuence that Zeisings ideas had
on classical music theory when one looks at the writings of Arnold
Schonberg [3], who wrote in his Harmonielehre (1911) in reference
to musical form:
I dont believe in the golden ratio. At least, I dont
believe that it is the single formal principle for our
sense of beauty; rather at most one among many,
among countless many.
The tone of this statement makes it clear that the idea of the golden
ratio had been discussed among music theorists for some time before
Schonberg wrote this.
Although composer Bela Bartok is not known to have discussed
his use of the golden ratio and Fibonacci numbers, it appears likely
that he consciously used them in his Music for Strings, Percussion
and Celesta (1936) [4]. Theorist Erno Lendvai [5] is credited with
this discovery. My own independent analysis has conrmed that
Bartok timed the climax of the rst movement to occur extremely
close to the golden section within about 1/1000 of the proportion,
depending on the tempi chosen (my own analysis used the average
A STYLE CHARACTERIZED BY FIBONACCI NUMBERS AND 3
of the recommended tempi). This theory is further supported by
Bartoks apparent use of a palindromic rhythm based on Fibonacci
numbers in the opening of the third movement:
It would be a great coincidence if Bartok did not consciously imple-
ment the sequence in this passage.
American composer Joseph Schillinger [6] suggested creating pitch
structures based on the Fibonacci sequence in his series of books
The Schillinger System of Musical Composition (1941). Iannis Xe-
nakis [7] used Fibonacci numbers in his Anastenaria, Le sacrice
(1953). Luigi Nono [8] used successive note-lengths derived from the
Fibonacci sequence in Il canto sospeso (1956). Gyorgy Ligeti [9] ac-
knowledged his use of the golden ratio in determining the lengths
of the respective sections in Apparitions (1958-59). Ernst Krenek
[10] composed a work entitled Fibonacci Mobile (1964), in which
he utilized the series in the construction of the proportions. Brian
Ferneyhough [11] writes of his application, in various roles, of the
proportions inherent in the Fibonacci series in his Time and Motion
Study I (1971-77). Karlheinz Stockhausen [12], in a 2004 interview
with Piergiorgio Odifreddi, stated that he used the Fibonacci num-
bers in many of his works, for the duration of dierent sections as
well as for the density of sound-masses.
2. The Origins of the Style
In the music of the composers mentioned above, the Fibonacci
numbers and golden ratio did not form the basis of a musical style.
These were isolated works, and in most of these composers works,
no trace of their use is evident.
In 1997, a great curiosity began to drive my work; I wanted to know
what a music would sound like in which Fibonacci numbers and the
golden ratio played the central role. In my rst compositions with
Fibonacci numbers, I used them as a means of providing structural
and tonal cohesion, similar to the composers discussed above. These
rst compositions were all written in 12-tone equal temperament
4 CASEY MONGOVEN
(using the notes of the piano), and in most of them, I used the
Fibonacci and Lucas sequences modulo 12. A few were written in
24-tone equal temperament using modulo 24. In 2001, I decided that
I was no longer interested in using a tuning system handed down to
me through tradition; after all, that tuning system had nothing to
do with Fibonacci. I set out to create my own tuning system.
3. A System of Tunings Based on Fibonacci and the
Golden Ratio
The base interval r in 12-tone equal temperament, the half-step,
is derived from
r =
12

2 1.059463 .
So if the note A on the piano is tuned to 440 Hz, that means the
next note up, A, is
440(
12

2) 466.2 Hz .
The cent serves as the logarithmic unit of measure for musical
intervals, and it is dened that
12

2 = 100 cents ;
there are therefore 1200 cents in one octave (the interval of an octave
is created by a 2:1 frequency ratio). What happens if we use the
golden ratio as our base interval instead?
r = =
1 +

5
2
833 cents
This is a rather large interval, lying approximately 33 cents higher
than the minor sixth. The human hearing range spans from approx-
imately 16 to 20000 Hz. How many pitches is it possible to have in
our hearing range using the golden ratio as a base interval?
1. 16 Hz
2. 25.9 Hz
.
.
.
15. 13488.0 Hz
If we start at 16 Hz, the highest note would be approximately
13488 Hz, and we would have a total of only 15 possible pitches.
This is an interesting tuning, but it is not very economical in terms of
A STYLE CHARACTERIZED BY FIBONACCI NUMBERS AND 5
pitch space. Let us try using the smaller golden ratio value = 1/
and
r =
2
+ 1 1.381966 560 cents .
How many notes are possible now?
1. 16 Hz
2. 22.1 Hz
.
.
.
23. 19727.0 Hz
This tuning also has rather large intervals. Let us create a system
of temperaments based on
r =
x
+ 1 ,
where x is a positive integer, noting that [13]
lim
n
F
n
+ F
nx
F
n
=
x
+ 1 .
Now we have a system of tunings based on the golden ratio and
Fibonacci numbers.
base interval approx. pitches in notes within one
size in cents hearing range one octave
r = = + 1 833 15 2
r =
2
+ 1 560 23 3
r =
3
+ 1 367 34 4
r =
4
+ 1 236 53 6
r =
5
+ 1 149 83 9
r =
6
+ 1 94 132 13
r =
7
+ 1 59 211 21
r =
8
+ 1 36 339 33
r =
9
+ 1 23 546 54
r =
10
+ 1 14 881 86
r =
11
+ 1 9 1423 139
r =
12
+ 1 5 2300 224
r =
13
+ 1 3 3719 362
r =
14
+ 1 2 6015 585
r =
15
+ 1 1 9731 946
6 CASEY MONGOVEN
These tunings contain no octaves or any other pure interval for
that matter.
4. Purification of the Style
Having created a system of temperaments, I decided that a style of
music characterized by Fibonacci numbers and the golden ratio had
to be comprehensible above all else. By this is meant that there could
not be any extra elements used in a composition that might obscure
the mathematical object which that composition was based on; it
had to be crystal clear how the music related to the mathematical
object it portrayed. Rather than becoming more complicated, as the
styles of most composers tend to do, my style became simpler in the
most fundamental sense. All elements of tradition in my music were
abandoned. I wrote the rst work in this style in April 2002, using
the software synthesis language Csound [14].
5. Characteristics of the Style, Compositional Methods
For each work in the style, a sequence related to the Fibonacci
numbers and the golden ratio is selected. Generally, only one se-
quence is used in a composition. Similarly, only one temperament
is used in a composition, selected from those described earlier. The
sequence is then graphed into musical space in a direct and simple
manner. By this is meant that each integer occurring in the sequence
is assigned a pitch; integers are often assigned other musical parame-
ters too, such as loudness, timbre and spatial location. In most cases,
the integers are heard one after another, ordered according to their
indices. To provide a simple example of what is meant here, let us
look at a graph of my work Horizontal Para-Fibonacci Sequence no.
31; the original score and sound les are available at [15]. This work
is based, as the name indicates, on the horizontal para-Fibonacci se-
quence (A035614 in The On-Line Encyclopedia of Integer Sequences
or OEIS [16]), which gives the column number in which integers
occur in the Wytho array.
A STYLE CHARACTERIZED BY FIBONACCI NUMBERS AND 7
Each point in the graph represents a musical note. Each note
receives an equal duration, in this case 50 milliseconds. This duration
is referred to as the note value. In this work, the temperament based
on
4
+1 was used. The lowest integer, 0, is given a dynamic value of
mezzopiano (medium-quiet); the highest integer is given a dynamic
level of forte (loud). The arrow pointing upwards in between these
two dynamic levels represents a gradual increase in loudness from
integers 0 to 12 (from mezzopiano to forte). 377 integers are used in
this composition a Fibonacci number. We can see from the graph
that every time a certain integer occurs in the sequence, it sounds
the same; this is critical for transforming the mathematical properties
into musical properties in a direct and perceptible manner. Every
time the integer 0 occurs in the sequence, the pitch heard is 205.7
Hz at a dynamic level of mezzopiano. The choice of temperament,
note value and dynamic level is a purely artistic decision.
If the pitch rises with the value of an integer, as in the previous
example, the orientation of the work is termed ascending. If the ori-
entation in a work is descending, the pitch falls as the value of the
integer increases. The following graph is from my piece S(n) Rep
Sequence no. 20; the sequence used here (A000119 in the OEIS) ap-
pears under that name in Fibonacci and Related Number Theoretical
Tables [17]. This sequence gives the number of possible represen-
tations of integers as sums of distinct elements of the Fibonacci se-
quence beginning 1, 2, 3, 5, 8, . . . . This is an example of descending
orientation. The original score and sound les are available at [18].
8 CASEY MONGOVEN
In this work, the temperament based on
5
+ 1 was used. The note
value is 58 milliseconds.
6. Notation
When I made the decision to abandon traditional elements of clas-
sical music and create my own system of tunings, it became clear that
classical notation would no longer be suitable for my style.
My notation has been developed and rened since I created the
rst works in this style. My current notation is written in HTML.
The graphs above are not from the original notation, and were cre-
ated specically for explanatory purposes. In the actual scores, the
parameters (meaning the pitches, dynamic level and other informa-
tion seen on the left here) are on a separate page from the graph,
and more information on the compositions is provided.
The following score has been slightly modied from its original
HTML form for explanation and practical purposes. The lists have
also been truncated (indicated by triple-dots) to save space. The
original score and sound les can be found at [19].
A STYLE CHARACTERIZED BY FIBONACCI NUMBERS AND 9
Collection VIII
Fibonacci Entry Points no. 24
Casey Mongoven
March 27, 2008
description of sequence: index of lowest Fibonacci number
that n divides
1
offset: 1
2
number of members used: 377
number of channels: 2
3
piece length: 20.358 seconds
note value: .054 seconds
orientation: descending
temperament: phi
12
+ 1
lowest frequency: 96.5591688 Hz highest frequency: 618.65
Hz
number of unique frequencies: 140
4
synthesis technique: wavetable
5
wave: wave one to wave two
6
wave one source: large balloon
7
wave two source: electric drill
8
dynamic: mp to f
simulated spatial location: right to left
9
delay: .0018 to -.0018
10
attack: .0063 to .0055 (roll-off .0000013)
11
release: .0102 to .0089 (roll-off .0000022)
synthesis language used: Csound
1) A001177 2) index of rst term used 3) a stereo work 4) number of
distinct integers (and therefore frequencies, loudness levels, etc.) in rst
377 members of this sequence 5) Wavetable synthesis involves using
recorded sounds and stretching them or compressing them to create
dierent pitches. 6) The timbre changes according to the sequence, just as
the loudness and pitch do, from wave one to wave two. 7) The sound of a
balloon letting out air was use as a sound source for wave one. 8) The
sound source of wave two was an electric drill. 9) Using psycho-acoustic
phenomena or stereo eect, each integer is given its own simulated
spatial location, so the spatial location of the sound is also governed by
the sequence. 10) Delay between the right and left speakers given in
seconds here is one of the elements used to create the spatial element.
11) Attack and release given in seconds refer to the time at the
10 CASEY MONGOVEN
beginning and end of a single tone, in which the sound gets louder and
quieter. Without attack and release, one would hear a click before and
after each note. The attack and release values are controlled by the
sequence here, so that higher integers sound crisper than lower ones.
int. frequency wave dyn. s.s.l. attack release
1 618.6500 w1*.78+w2*.22 mp right .006300 .010200
3 614.8252 .006297 .010195
4 612.9217 .006295 .010193
5 611.0241 .006294 .010191
6 609.1324 .006293 .010189
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
500 131.6618 .005633 .009117
534 118.4877 .005588 .009043
600 96.55917 w1*.21+w2*.79 f left .005500 .008900
sequence values used: 1, 3, 4, 6, 5, 12, 8, 6, 12, 15, 10, 12,
7, 24, 20, 12, 9, 12, 18, 30, 8, 30, 24, 12, 25, 21, 36, 24,
14, 60, 30, 24, 20, 9, ..., 72, 534, 358, 60, 342, 90, 220,
168, 185, 60, 368, 24, 60, 285, 216, 60, 187, 90, 500, 48, 14
graph:
A STYLE CHARACTERIZED BY FIBONACCI NUMBERS AND 11
7. Multimedia Works
At the end of 2007, I realized a goal I had had for some time to
create a multimedia work in this style. Using the graphic program-
ming language Processing, I created a presentation of 13 works in
which point graphs were showed synchronized with the music. This
means that when a tone is played, it appears on the screen at exactly
that moment. A somewhat minimalist style was used for the graphic
element; the tones appeared in white on a black background while
they were being played, fading to grey once they had been heard.
The response from the audience was quite positive, and I found that
people had a better understanding of what was happening in the
music.
8. Concluding Words
With this music, I hope to answer the age-old question of what a
musical style sounds like which has its roots in the Fibonacci num-
bers and the golden ratio. Many more numeric sequences related
to mathematics of Fibonacci numbers and golden ratio are used in
my music. These include among others; Zeckendorf representations
(A014417), Wytho and Stolarsky arrays [20], base Phi [21] and base
Lucas representations (A130310 and A130311), signature sequences
based on Phi (A084532, A084531 and A118276), and Pisano periods
(A001175). Catalogue entries on my website always contain sound
les, scores and links to the sequence used in the OEIS. To hear
more works in this style, visit http://caseymongoven.com; I oer the
public free access to my music.
Special thanks goes to Ron Knott for his support over the years
and comments while I was writing this paper.
References
[1] Adolf Zeising. Das Normalverhaltniss der chemischen und morphologischen
Proportionen. Leipzig: Rudolph Weigel, 1856.
[2] Gustav Theodor Fechner. Zur experimentalen

Asthetik. Leipzig: Hirzel,
1871.
[3] Arnold Sch onberg. Harmonielehre. Wien: Universal Edition, 2001.
[4] B ala Bart ok. Music for String Instruments, Percussion and Celesta. New
York: Boosey & Hawkes, 1964.
[5] Ern o Lendvai. Bart oks Style: As Reected in Sonata for Two Pianos and
Percussion and Music for Strings, Percussion and Celesta. Translated by
Paul Merrick and Judit Pokoly. Budapest: Akkord Music Publishers, 1999.
12 CASEY MONGOVEN
[6] Joseph Schillinger. The Schillinger System of Musical Composition. New
York: Carl Fischer, reprint 1946.
[7] Iannis Xenakis. Preface to Anastenaria, Le sacrice. Paris: Salabert, 1953.
[8] Jonathan Kramer. The Fibonacci Series in Twentieth-Century Music,
Journal of Music Theory, vol. 17, no. 1 (spring 1973): pp. 110-148.
[9] Peter Varnai. Gy orgy Ligeti in Conversation: with Peter V arnai, Josef
H ausler, Claude Samuel and Himself. London: Eulenberg, 1983.
[10] Will Ogdon, Ernst Krenek. Conversation with Ernst Krenek, Perspectives
of New Music, vol. 10, no. 2 (spring - summer 1972): pp. 102-110.
[11] Brian Ferneyhough. 2001. Program notes from Concert XX,
CONCERTEN Tot en Met. http://concerten.free.fr/m/c/program/XX.html
(accessed April 2008).
[12] Piergiorgio Odifreddi. 2004. Intervista a KARLHEINZ STOCKHAUSEN,
vialattea.net.
http://www.vialattea.net/odifreddi/stockhausen/stockhausen.html
(accessed April 2008).
[13] Steven Vadja. Fibonacci and Lucas Numbers, and the Golden Section.
Mineola: Dover, 1989.
[14] Cesare Marilungo, Richard Boulanger. cSounds.com. http://csounds.com
(accessed April 2008).
[15] Casey Mongoven. 2008. Mongoven B841.
http://caseymongoven.com/catalogue/b841.
[16] Neil Sloane. 2007. The On-Line Encyclopedia of Integer Sequences.
http://www.research.att.com/njas/sequences/.
[17] Brother Alfred Brousseau. Fibonacci and Related Number Theoretical
Tables. Santa Clara: The Fibonacci Association, 1972.
[18] Casey Mongoven. 2008. Mongoven B843.
http://caseymongoven.com/catalogue/b843.
[19] Casey Mongoven. 2008. Mongoven B823.
http://caseymongoven.com/catalogue/b823.
[20] David R. Morrison. A Stolarsky Array of Wytho Pairs, A Collection of
Manuscripts Related to the Fibonacci Sequence. Edited by Verner E.
Hoggatt, Jr. and Marjorie Bicknell-Johnson. Santa Clara: The Fibonacci
Association, 1980.
[21] Ron Knott. 2007. Phigits and the Base Phi Representation, Fibonacci
Numbers, the Golden Section and the Golden String.
http://www.mcs.surrey.ac.uk/Personal/R.Knott/Fibonacci/ phigits.html
(accessed April 2008).
AMS Classication Number: 11B39
Jakobstrae 3, 99423 Weimar, Germany
E-mail address: cm@caseymongoven.com